Patentable/Patents/US-11488015
US-11488015

Artificial intelligence layer-based process extraction for robotic process automation

PublishedNovember 1, 2022
Assigneenot available in USPTO data we have
Inventorsnot available in USPTO data we have
Technical Abstract

Artificial intelligence (AI) layer-based process extraction for robotic process automation (RPA) is disclosed. Data collected by RPA robots and/or other sources may be analyzed to identify patterns that can be used to suggest or automatically generate RPA workflows. These AI layers may be used to recognize patterns of user or business system processes contained therein. Each AI layer may “sense” different characteristics in the data and be used individually or in concert with other AI layers to suggest RPA workflows.

Patent Claims
13 claims

Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.

Claim 2

Original Legal Text

2. The computer-implemented method of claim 1, wherein the data collected by the listeners comprises user interactions with respective user computing systems, audio, video, light, heat, motion, acceleration, radiation, or any combination thereof.

Plain English translation pending...
Claim 3

Original Legal Text

3. The computer-implemented method of claim 1, wherein the multiple AI layers comprise a sequence extraction layer, a clustering detection later, a visual component detection layer, a text recognition layer, an audio-to-text translation layer, or any combination thereof.

Plain English translation pending...
Claim 4

Original Legal Text

4. The computer-implemented method of claim 1, wherein each individual AI layer has an associated modifier based on an estimated accuracy of the respective individual AI layer.

Plain English Translation

This invention relates to a computer-implemented method for improving the performance of artificial intelligence (AI) systems by dynamically adjusting the contributions of individual AI layers based on their estimated accuracy. The problem addressed is the variability in performance across different layers of an AI model, where some layers may be more reliable than others. The solution involves assigning a modifier to each AI layer, which scales its output based on its estimated accuracy. Higher accuracy layers receive greater influence in the final decision, while lower accuracy layers are downweighted. The method ensures that the AI system leverages the most reliable components, improving overall accuracy and robustness. The modifiers can be updated dynamically as the AI system processes new data, allowing the system to adapt to changing conditions. This approach is particularly useful in complex AI models where different layers may specialize in different aspects of the problem, and their reliability may vary depending on the input data. By dynamically adjusting the influence of each layer, the system can achieve better performance than a static model where all layers contribute equally. The invention can be applied to various AI applications, including image recognition, natural language processing, and decision-making systems.

Claim 5

Original Legal Text

5. The computer-implemented method of claim 1, wherein an RPA workflow is only generated when a collective confidence threshold for all AI layers has been exceeded.

Plain English Translation

This invention relates to automated workflow generation using robotic process automation (RPA) and artificial intelligence (AI). The problem addressed is ensuring that RPA workflows are only generated when there is sufficient confidence in the AI-driven decision-making process, preventing errors or unreliable automation. The method involves multiple AI layers analyzing data to determine whether an RPA workflow should be created. Each AI layer evaluates different aspects of the process, such as data accuracy, task feasibility, and potential risks. The system calculates a collective confidence score based on the outputs of all AI layers. Only when this collective confidence score exceeds a predefined threshold is the RPA workflow generated. This ensures that automation proceeds only when the AI system is highly confident in its recommendations, reducing the risk of errors in the automated workflow. The AI layers may include machine learning models, rule-based systems, or other decision-making algorithms. The collective confidence threshold is set to a level that balances automation efficiency with reliability, ensuring that only well-supported workflows are generated. This approach improves the reliability of RPA systems by incorporating AI-driven validation before automation is executed.

Claim 6

Original Legal Text

6. The computer-implemented method of claim 1, wherein the multiple AI layers are configured to perform statistical modeling and utilize deep learning techniques to identify the one or more RPA processes in the collected that exceed a confidence threshold.

Plain English translation pending...
Claim 8

Original Legal Text

8. The computer-implemented method of claim 7, wherein similarity between the existing process and the identified RPA process is determined by entropy, minimization of a process detection objective function, or a combination thereof.

Plain English translation pending...
Claim 9

Original Legal Text

9. The computer-implemented method of claim 1, wherein the collected data is run through the multiple AI layers in series.

Plain English Translation

This invention relates to a computer-implemented method for processing data using multiple artificial intelligence (AI) layers arranged in series. The method addresses the challenge of efficiently analyzing large datasets by leveraging sequential AI processing to enhance accuracy and performance. The system first collects data from one or more sources, which may include sensors, databases, or user inputs. The collected data is then processed through a series of AI layers, each designed to perform a specific analytical task. These layers may include machine learning models, neural networks, or other AI algorithms, each configured to extract different types of insights or features from the data. By processing the data sequentially through these layers, the method ensures that each layer builds upon the outputs of the previous one, improving the overall accuracy and efficiency of the analysis. The method may also include steps for validating the processed data, ensuring that the results meet predefined quality standards before being used for decision-making or further applications. This approach is particularly useful in fields such as predictive analytics, autonomous systems, and real-time data processing, where accurate and timely insights are critical.

Claim 10

Original Legal Text

10. The computer-implemented method of claim 1, wherein the collected data is run through the multiple AI layers in parallel.

Plain English translation pending...
Claim 11

Original Legal Text

11. The computer-implemented method of claim 1, wherein the collected data is fed through a combination of both series and parallel AI layers.

Plain English translation pending...
Claim 15

Original Legal Text

15. The non-transitory computer-readable medium of claim 14, wherein similarity between the existing process and the identified RPA process is determined by entropy, minimization of a process detection objective function, or a combination thereof.

Plain English translation pending...
Claim 16

Original Legal Text

16. The non-transitory computer-readable medium of claim 12, wherein the data collected by the plurality of listeners comprises user interactions with respective user computing systems, audio, video, light, heat, motion, acceleration, radiation, or any combination thereof.

Plain English Translation

This invention relates to a system for collecting and processing diverse types of data from multiple sources, particularly in environments where real-time monitoring and analysis are required. The system includes a plurality of listeners, each configured to gather data from various sources such as user interactions with computing systems, audio, video, light, heat, motion, acceleration, and radiation. These listeners operate in a distributed manner, allowing for scalable data collection across different environments. The collected data is then transmitted to a central processing unit, which processes the data to extract meaningful insights. The system is designed to handle heterogeneous data types, enabling comprehensive monitoring and analysis in applications such as smart environments, industrial automation, or security systems. The invention addresses the challenge of integrating and processing diverse data streams efficiently, providing a unified framework for real-time data analysis. The system ensures robustness by validating and filtering the collected data before processing, reducing noise and improving accuracy. This approach enhances decision-making processes by providing timely and reliable data-driven insights.

Claim 17

Original Legal Text

17. The non-transitory computer-readable medium of claim 12, wherein the collected data is run through the multiple AI layers in series or the collected data is run through the multiple AI layers in parallel.

Plain English Translation

This invention relates to a system for processing data using multiple artificial intelligence (AI) layers, addressing the challenge of efficiently analyzing large datasets with varying computational demands. The system collects data from one or more sources and processes it through a series of AI layers, which may be arranged in either a serial or parallel configuration. In the serial arrangement, data flows sequentially through each AI layer, allowing for progressive refinement or hierarchical analysis. In the parallel arrangement, data is processed simultaneously across multiple AI layers, enabling faster processing or parallelized feature extraction. The AI layers may include different types of neural networks, machine learning models, or other AI techniques, depending on the specific application. The system dynamically selects the processing mode (serial or parallel) based on factors such as data volume, computational resources, or processing time constraints. This adaptability ensures optimal performance across different scenarios, from real-time analytics to batch processing. The invention improves efficiency by leveraging the strengths of both serial and parallel processing architectures, making it suitable for applications in data analysis, predictive modeling, and automated decision-making.

Claim 18

Original Legal Text

18. The non-transitory computer-readable medium of claim 12, wherein the collected data is fed through a combination of both series and parallel AI layers.

Plain English translation pending...
Classification Codes (CPC)

Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.

Patent Metadata

Filing Date

December 9, 2019

Publication Date

November 1, 2022

Want to explore more patents?

Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.

Citation & reuse

Analysis on this page is generated by Patentable — an AI-powered patent intelligence platform. AI-generated summaries, explanations, FAQs, and analysis may be reused with attribution and a visible link back to the canonical URL below. Patent abstracts and claims are USPTO public domain.

Cite as: Patentable. “Artificial intelligence layer-based process extraction for robotic process automation” (US-11488015). https://patentable.app/patents/US-11488015

© 2026 Nomic Interactive Technology LLC. Machine-readable context available at /api/llm-context/US-11488015. See llms.txt for full attribution policy.